Pinned Loading

1. nyc-taxi-data-engineering-postgresql nyc-taxi-data-engineering-postgresql Public

   Scalable ETL pipeline and PostgreSQL optimization for 40M+ records. Features memory-efficient Python preprocessing and B-Tree indexing for a 95% reduction in query latency.

   Python

2. bangladesh-dengue-forecasting bangladesh-dengue-forecasting Public

   A 15-year longitudinal study evaluating climate covariates in national-scale dengue forecasting. Contrasts SARIMA and Random Forest models under strict rolling-origin validation.

   Python

3. Bangladesh-air-quality-forecasting Bangladesh-air-quality-forecasting Public

   Reproducible time-series pipeline for PM2.5 forecasting across 8 cities in Bangladesh. Implements automated data ingestion (Open-Meteo API) and ARIMA/SARIMA models with rolling-origin validation.

   Python

4. Clean-room-ml-regression Clean-room-ml-regression Public

   NumPy-based implementation of Linear and Logistic Regression from first principles with vectorized gradient descent, regularization, convergence diagnostics, and empirical validation.

   Python

5. FRED-CPI-ARIMA-forecasting FRED-CPI-ARIMA-forecasting Public

   CPI forecasting using ARIMA(5,1,0) with 12-month out-of-sample validation and benchmark comparison against a naive model.

   Python

6. dengue-forecasting-system-deployment dengue-forecasting-system-deployment Public

   Python